封面
声明
致谢
中文摘要
英文摘要
目录
变量注释表
1 绪论
1.1研究背景与意义(Research Background and Basic for Selected Topic)
1.2相关领域的研究动态(Overview of the Related Research)
1.3研究内容和技术路线(Research Contents and Technical Flow)
1.4论文组织结构(Structure of Thesis)
2 多元逻辑回归高光谱遥感影像监督分类
2.1多元逻辑回归(Multinomial Logistic Regression)
2.2回归参数求解策略(The Solution Strategy Of Regression Parameter)
2.3 DFP修正拟牛顿算法(DFP Quasi-newton)
2.4本章实验数据(Experiment Data)
2.5实验结果与分析(Experiment Results and Analysis)
2.6本章小结(Brief Summary of This Chapter)
3基于多元逻辑回归分类器的半监督样本选择
3.1非标记样本选择(The Selected Unlabeled Samples)
3.2常用样本增选策略(Original Algorithms of Sample Select)
3.3基于选择性方差的样本增选方法(A Sample Select Algorithm Based on Selectivity Variance )
3.4结果与分析(Experiment results and Analysis)
3.5本章小结(Brief Summary of This Chapter)
4基于邻域信息和多分类器的半监督标签确定
4.1非标记样本标注分析(Analysis of Labeling Unlabeled Sample)
4.2常用非标记样本的确定方法(Original Algorithms of Label Sample)
4.3基于邻域信息和多分类器融合的遥感半监督分类(A Semi supervised Algorithm of Label Sample Based on Neighborhood Information and Classifier Combination )
4.4实验结果与分析(Experiments and Analysis)
4.5本章小结(Brief Summary of This Chapter)
5结论与展望
5.1 结论(Conclusions)
5.2 展望(Prospects)
参考文献
作者简历
学位论文数据集